IMPACT I N REAL STATE VALUE OF URBAN AND

Anuncio
I MP A CT
IN
REAL
S TA T E V A L U E
OF
URBAN
AND
TR A N S P O RT AT I O N
P R O J E CT S , I N T H E M ET R OP OL I T A N A R E A O F B A R C E LO N A
1
2
Jorge Cerda Troncoso , Claudia Pérez Prieto y Carlos Marmolejo Duarte
3
Cent ro de Política de Suelo y Valoraciones
Universidad Polit écnica de Cataluña
A BST RACT : Oft en, research has studied the urban impact through: urban development and
transport projects. The results suggest that factors from the real estate products (surface
useful, number of bedrooms and bat hrooms, parking, swimming pools, air conditioning, age)
and environmental factors (socioec onomic status of the resident population, the proximity to
infrastructure road and rail, the diversity of ec onomic activities) are included in the expectation
of real estate value. In this paper we report the results of a study that tries to simulate an
integrated manner, the effect produced by large urban interventions on the accessibility, land
use and real estate value in the Metropolitan Area of Barcelona. Using integrated transport
model (model double bounded distribution and network assignment model) and hedonic models
in a specialized GIS platform will try to simulate at census tract scale, direct impacts in the area
of implement ation of a specific development project, and also the indirect impact that o ccurs in
the system of interaction of the metropolitan area of Barcelona. The analysis suggests that each
group of urban development (approximately 30), produc e different effects on the dimensions
analyzed. The effect of transport projects is significantly lower than the effect of urban
interventions as a whole, in the value of s upply. This is because urban interventions, in addition
to reconfigure the structure of accessibility, change the intensity of the attractions and
generations of the system, a fact that no occur with only the transportation projects. There are
cases where the magnitude of urban development is significant; however t here are no changes
in property values.
Palabras clave: impact urban, urban development, transportation projects.
1
Contact Author: [email protected]
Contact Author: [email protected]
3
Contact Author: [email protected]
Centro de Política de Suelo y Valoraciones, www.upc.es/cpsv
Universidad Politécnica de Cataluña.
C/ Jordi Girona s/n, Local La CUP, Nord UPC. Edifici C3-C4. C. Postal 08034, Barcelona.
2
1. I NT RO DUCT I O N
The large number of projects that are scheduled to be developed in the urban area of Barcelona
originated the research question from this paper, referring to estimate the effect or urban impact
produced by t he realization of large urban i nterventions and transportation projects in the
dimensions of accessibility, land use and real estate value in the metropolitan area of
Barcelona. The ability to assess and in some cases simulate multidimensional impact of these
urban projects, is increasingly feasible due to the high sophistication of models of specific
topics, such as transport models and / or hedonic pric e models generate a new platform for
virtual urban management, multidimensional, with great potential.
The impact of transport projects is part of traditional strategic studies. Strategic level groups of
projects are evaluated according to current urban conditions, and future of the areas
considered. The methodology traditionally applied in these assessments, is called the classical
model of transport. As for the parameters of evaluation of a transportation project, has been the
effect on journey times for vehicles and fuel consumption in the whole metropolitan system
(ATM, 2007). Other research has incorporated additional variables to evaluate projects such as
emission of gaseous pollutants, accident conditions of transport net works, noise pollution,
generation of vibration, etc. (IDB, 2006).
Conc erning the assessment of the impact on property values, there is an extensive literature
which is mainly based on the methodology for the use of hedonic models. This method,
belonging to the family of revealed preference assumes that the value of assets implies the
marginal value of its attributes. In practice, it uses the property value (rent, buy, soil, etc.) of
buildings, to infer, econometrically, the marginal value of externalities, once the rest of locative
and manmade attributes have been controlled. The main strength of the method is that infers
the implicit price of environmental attributes from the actual behavior of individuals in the
market.
According Sirmans et al. (2006), the proper factors of real estate products were similar in the
sense of identifying a signific ant effect of the conditions of surface useful, lot size, number of
bedrooms and bathrooms, parking, swimming pools and air conditioning. Another int rinsic
characteristic of these products and relevant effects is the age of the building. Inc reasing age of
the construction inc reases the possibility of a renewal, thereby increasing the property value
expectations. The evolution of this situation indicates that up to 27 years the property suffers the
effect of the depreciation (reaching 53% of initial value). Then the possibility of renewed
increases in value by 32%, bringing the price up to 87% off the original price (Song et al., 2005)
With regard to urban environmental factors were similar in t he sense of identifying a value enhancing effects of some conditions of t he urban environment such as socioeconomic status
of the resident population (a higher level of income greater value), the proximity to infrastructure
road and rail (De Sousa et al., 2009), the diversity of economic activities associated with the
home and neighborhood level, as the increased scale of business and servic es diminish the
value of homes (congestion, noise and ins ecurity.) (Song and Knaap, 2004) In referring to the
neighborhood level (Simons et al., 1998) reported t wo factors, the c onstruction of new homes
and delinquency in the payment of property taxes, which affect the sale price of nearby
properties. The latter, an indicator is seen as an indicator of disinvestment by the neighboring
property, and verifies a negative impact on prices. In cont rast, new construction, despite having
a significant and positive effect on prices, the degree of spatial concentration is, make a
differenc e in terms of the importance of this effect. Thus, the effects are magnified near the big
new projects, which could have implications for public investment, in the sense to maxi mize the
efficiency of the positive externalities of housing subsidy for larger projects.
Regarding environmental factors, Hiteb (2005) shows a low effect of environmental pollution in
the price of housing. But in addition this study constructs a demand c urve for environmental
conditions from the hedonic model. This demand model is verified that the most demanding
environmental quality is the upper class, highly educated, and families with children. Curve is
relatively inelastic. They fit (same behavior) the demand for educational quality of the
environment with t he demand for environmental quality. The size of dwelling and environmental
quality, are substitute goods, while t he size of the lot turns out to be complementary to the
environmental quality.
The s ocial dimensions and economic activities don’t have developments so clear, as those
mentioned above, in relation to the impact assessment approach. Under a different conceptual
approach, have been the transport models that have c ome over, in this sens e, the development
of so-called land use models and transportation (LUTM), also known as five-stage models
(Hunt, JD et al., 2005). These models predict the location of activities (residential, economic) at
the level of an entire urban conglomeration, conside ring a sequenc e of effects on soil attributes,
and the transport network, dynamically over time. That is, the location depends on the real
estate market conditions, which depend in part of the attributes of urban areas, one of which is
the level of accessibility, which in turn is affected by the location of activities. In this sens e,
Gibbons and Machin, (2008), suggests that there is no consensus on the influenc e of changes
in transport prices, which is not surprising, since the utility of new infrastructur e or other
changes in transport policy depends largely on economic opportunities in terms of accessibility,
and the overall urban context.
In this paper, we studied the impact (through an econometric analysis) on accessibility, land use
and property values, large urban developments and t rans port projects planned for the area east
and west of the central conurbation B arcelona. The rest of the paper is structured as follows:
first define objectives and area of study, then explain the methodology and data us ed, and
finally discuss the results of data analysis.
2. O BJ ECT IV ES O F T HE ST UDY
In this research the main objective is to analyze the impact on accessibility, land use and
property values, of large urban development (with final approval or advanced study in process),
and t ransportation projects listed in Plan for Transport Infrastructure (PDI), planned for the area
east and west of the central conurbation of B arcelona. The specific objectives are intended to 1)
construct a hedonic pric e model, to evaluate the effect of current conditions in the area
(accessibility, morphology, and socio-professional structures) on the spatial configuration of
property values (offers for sale and rental) residences, offices, retail shop, 2) quantify the impact
in physical terms (surface land and roof spac e), ec onomic (employment), demographic
(households and individuals), and transport (accessibility), of new development from planning ,
potential project without a development plan and the projects considered in the master plan f or
transport infrastructure, 3) evaluate the effect of the new urban configuration (resulting from
implementation of new development, and transport) in real estate values in the project area, and
the rest of the territory, and 4) make a territorial hierarc hy based on the differential effect (with
and without projects) in urban and real estate value attributes.
3. D AT A A ND M ET HO DOL O GY
The overall objective of the study foc uses on t he central conurbation of the city of B arcelona,
with an area of 370 km2, and c onsists of 23 municipalities, 2382 census areas, 351 areas of
transport, and 5 protosystem. Figure 1, shows the study area arising from the overlap of
functional protosystem (areas grouping several municipalities in terms of its structure interaction
of travel) municipal boundaries, cens us tracts, and t he continuous urbanized (obtained by the
semi-assisted proc essing of satellite images SPOT). Solid black color shows the large urban
developments that are considered in the analysis.
Figure 1 Area of study
Source: Own CPSV
With regard to the Metropolitan Region of Barcelona (hereaft er RMB), the chos en area of study
represents 60% of the population, 59% of jobs, 33% of the buildings (where a higher proportion
of buildings in poor condition) and 66% of retail shop. From the above we can see the
significant weight that has defined the area of study, being the area of greatest density, diversity
and conc entration of bot h economic and demographic activity.
The impact of large urban devel opment and transport infrastructure projects are studied at two
scales. The scale of urban developments is census tracts and transport projects to scale
transport areas. For information on population and housing and buildings used 2001 data from
the National Statistical Institute of Spain. As information relating to travel, we use the survey of
daily travel (EMQ) 2001 of t he RMB, and travel times of t ravel by public transport and private
transport comes from the Autoridad Met ropolitana del Transporte (A TM). For its part, the
urbanized area is t he set of artificial land uses derived from analysis of remot e sensing from
satellite images of t he Corine Land Cover 1990 and 2000. For the analysis of the road net work
with its various hierarchies of pathways (associated with service levels, capacity, etc.), in the
RMB has used the information available for Teleatlas 2004. Finally information on property
values (residential, office and retail) comes from assessments and offer prices generated and
analyzed by studies of the Cent er for Land Policy and valuations (hereafter CPSV).
3.1 Urban projects to consider
For the selection of large urban developments, we used the information obtained by the study
“El Potencial Urbanístic de la Regió Metropolitana de Barcelona, una visió des del 2007 “,
conducted by the CPSV in 2007. According to the study area, 169 projects were selected from a
total of 1,262 projects that considers the Urban Potential Study of the RMB.
In terms of roof space, the area of study includes projects for a total of 24.66 million m2,
representing 27.5 % of the total planned for the RMB (91.08 million m2 of roof space). Most are
intended for economic activity, followed by residency, underground parking, and finally
equipments. In terms of spatial distribution, this area is locat ed mainly in the municipalities of:
Barcelona t o 12.13 million m2 roof space, El Prat de Llobregat to 4. 06 million m2 roof space
and L'Hospitalet de Llobregat 1.89 m2 million roof spac e.
The overall roof space in the area of study has the capacity to serve 324. 201 workplaces,
corresponding to 47% of the t otal planned for the RMB. This percentage is significant, because
in terms of roof spac e, the area of study focus only 27% of the RMB, which is symptomatic of
the greater density and tertiary level of economic activity that is expected to be installed in this
area.
Of a tot al of 169 projects planned for the study area were grouped bas ed on their location in the
area of study, forming 30 groups presented in Table 1.
Table 1 List of groups.
Source: Own CPSV
3.2 Transport projects to consi der
For the selection of transportation projects use t he information derived from the Master Plan for
Trans port Infrastructure (P DI) 2001-2010 finally approved on April 25, 2002 by board of A TM.
The P DI 2001-2010, considered a total of 19 projects to expand the rail network, 16 projects of
modernization and improvement of net works (Metro and FGC), 11 projects of heat exchangers,
19 projects of expansion, modernization and improvement of the rail net work state, and 12
feasibility studies and design of new projects. The year 2007 is held the fourth annual
monitoring report of the P DI, whic h makes known the status of implementation of each of the
projects approved, and the state of feasibility studies.
Figure 2 shows the location of the projects considered in 2026 and the network bas e (2007).
Figure 2 Transport network baseline, scenario 2007, and projects PDI considered in 2026.
Source: Own CPSV
Source: Own CPSV
Subway projects are most important because it means an increase of 70% of the original
network (79.757 Km). These projects include the new line 9 which connects the Zona Franc a La
Sagrera, creating a corridor orbital public trans port, non-existent to date, in the city of
Barcelona. Next in importance are the Tram projects, which increase the original network in
31% (9.173 km), and finally the FFGC projects (train) which increase in 15% (3.338 Km.) the
original network.
3.3 Methodology
This research is important to achieve the integration of data from the base situation and the
data delivered by the projects. To achieve integration, it is necessary to conduct an analysis of
thematic and spatial coherence, namely it tries to find the variables of impact of projects that are
similar to thos e of the t errit orial baseline information, for this to be able to add them (if the
project's increase), or subt ract if the reduc ed project. The projects provide information
concerning: economic activities, housing, and work places located. This requires classifying the
land base information in three dimensions: characteristics of the park built, socioeconomic
structure of the residents, and structure of economic activities.
In the case of buildings, land base variables are: the age expressed by the weight ed average
year of construction and cons ervation status, classified into good, poor, bad and dilapidated.
Projects will int egrate these variables with t he total number of homes converted to number of
buildings by means of an index of buildings. Thus, the number of buildings from projects is
introduced to calculate the year of construction for the 2007 and in a good condition.
To integrate data from the socioeconomic structure of the residents use the housing from
projects which are divided int o three groups: HPO, concerted and free, and Resident Employed
Population (POR), separated in the 10 categories.
To integrate data from the structure of economic activities are used LTL (jobs) disaggregated
data in the same 10 categories that were grouped the POR. These categories are re-grouped in
a first group corresponding t o the tables high (managers, technicians and professionals,
technicians and support professionals and administrative staff), a second group c orresponding
to the middle management (services and sellers) and a third group tables corresponding low
(skilled workers, operators and unskilled workers. Finally the last group corresponds to the
armed forces and qualified agricultural workers. These grou ps capture the information that
determines the potential LTL as urban activities. Whereupon, the number of LTL offices,
technology park and equipment are added to the high tables, those derived from tertiary uses,
commercial, hotels, equipment, technical services and parking are added to the middle and
finally the LTL from the uses of Industry and logistics add to the low tables.
To characterize the socioeconomic structure of the residents used the information from the
resident employed population (hereafter POR), according to occupational category (executives,
professionals, technicians, salesmen, skilled workers, unskilled, etc.). In order to synthesize this
information applies the statistical technique of principal factor extraction where we get two
components. The first with a variance of 53%, has a positive value when grouping the
occupations of low income level (workers, skilled and unskilled workers ), and a negative value
when grouping the occupations of high income level (managers, professionals, and
technicians). This new variable is called “low- high level socioeconomic factors ", because on
the one hand polarized high income groups, and other low-income. With a variance of 13%, the
second component comprises administrative, and a smaller percentage of the technicians and
salesmen. The positive value factor is associated to predominance of administration and the
negative value when t hose are not dominant. This new variable is called "mid-level
socioeconomic factors”.
To characterize the externalities of economic activities, it uses information from the LTL under
the same occupational categories used in the socioeconomic analysis of the residents. But in
this case, different categories were grouped to achieve c onsistency with information from urban
development projects. LTL groups are: 1) managers, professionals, and administrative, 2)
Service and sales workers and 3) Operators skilled and unskilled. In order to synthesize this
information is made a factor analysis of the located workplace (using the varimax rotation
method) where you get two components. With a 66.5% variance, the first component compris es
professional and managers overall with skilled and unskilled operators. This new variable is
called "industrial activity factor." The positive value of this variable indicates a predominance
skilled and unskilled operator, while a negative value indic ates a predominanc e of managers
and professionals.
The functional interaction is defined as the level of attractiveness of the t errit ory and the
standard of accessibility that characterizes it, depending on the service levels of different
transport networks. To characterize eac h territory construct two indicators, which are also
directly altered with the implement ation of both urban projects, like transportation projects. The
first indicator corresponds to the attractiveness of travel to shop and reflects the travel density of
arriving in a territory with the purpose of shopping. The second indicat or is accessibility, for
which we must consider that the implementation of a development project will affect, first, the
structure of mobility needs in origin (residential), and secondly, the supply of work place and
commercial activities in the destinations.
In this context, what matters is how it will benefit the travel time of employees to their
work places, and the travel time of residents to shopping areas. As the travel time is a variable
that has a probability distribution is calculated the expected value, or rather, the most likely time
of travel. That is, each territory was estimated the most likely time to access a workplace
(Weighted Time in origin, hereafter TPO), and als o the most likely time to get to that territory
(Weighted Time in destination, hereaft er TP D). The same applies to travel shop, being in t his
case the TP O the most likely time to access a shopping center, and t he TP D t he most likely
time to reach a shopping center.
Tij *Vij
TPOi
j
Vij
j
Tij *Vij
; TPDj
i
Vij
i
(1)
To calculate the time weighted (Eq. 1) of each territory requires a matrix of travel (Vij), and a tra vel time
matrix (Tij) between each pair of zones i and j.
To analyzed real estate ex ogenous and endogenous factors. Among the exogenous factors are
considered variables that account for the morphology of the built environment. The variables
considered are: 1) average year of construction in the census tract, and 2) the number and
density of dwellings, surface range, considering three ranges three ranges from 0 to 75 m2,
from 76 to 90 m2, and over 91 m2. Endogenous factors refer to the characteristics of real estate
product analyzed. 1) The variables considered for residence (sale and rental) are: number of
bedrooms, number of bathrooms, floor area, holding a lift, quality bathrooms, quality kitchen,
old, and year of reform. 2) The variables considered for office (rental): total quality, which is
variable in the study itself that causes the database, and integrates a number of qualities
evaluated for different elements (roofs, walls, floor, etc.). And finally, 3) the variables for retail
(sales and rentals) are: axis shopping type nearest, distance from t he center of the nearest axis
shopping, total floor area, floor area at the level of access, front and bottom lengths of retail
shop.
Hedonic Pricing Model
Using the hedonic pricing met hod attempts to deduce the effect of different variables on a
decision or observed attribute. The effects of each variable are represented by the coefficients
of the econometric model that relates variables with attributes, or elections.
Regarding the real estate value, the method assumes that the value can be decomposed into
the marginal value provided by eac h one of the elements that characterize it. Thus, the property
value or offer price (P) is the integration of the value of each covariate (explanatory variable
multiplied by its shadow price, or price hidden).
Pi exp
o
i
i Xi
ln( Pi )
o
iXi
(2)
i
Where: Pi: i bid price, exp: exponential function, βo, βi: regression coefficients of each of the variables (Xi)
considered, Xi: exogenous and endogenous variables supply.
The functional expression used (Eq. 2) in this case is a semilog structure, where the
conventional calibration method, mostly used in the lit erature, goes through the application of
the natural logarithm function, then a linear regression calibration. Besides this func tional
structure has the advantage of determining (in the regression coefficient) directly the semielasticity, namely the impact on property values in percentage terms of unit changes in the
covariates.
In this way calibrated models to explain the total value and unit value (per m2 roof space) of the
property analyzed.
The strategy for calibration of the models are not focused to get the best adjust (highest level of
explanation) the real estate value, but rather the calibration coefficients that were c orrect (with
attention to basic assumptions of ec onometric techniques, which ensure efficient coefficients).
In order to eliminate the error of the econometric model, we performed a differential analysis
with and without project situations, thus providing the isolated effect of the implement ation of the
project, in the property value structure and the basic conditions of the territory.
Future status of the transport system
To assess the impact of different urban development in the territory, was necessary first to
isolate the effect of t heir actions or projects planned transport ation system. The configuration of
the transport system directly affects the time -weighted calc ulations in origin and destination (for
work and shopping), variables that ultimately are incorporated into the hedonic model.
Finally, keeping the matrix of travel from t he base year are calculated new values of the
indicators of accessibility, incorporating the impact on travel times of planned projects (PDI).
The result of this procedure shows that the greatest benefits will be generat ed in t he superior
orbital corridor of the city of Barcelona, bet ween the area of L'Hospitalet and P rat, mainly due to
commissioning of the new metro line 9. The greatest benefit means reductions of approxi mat ely
1 to 2 minut es of TPO to work (home work). In this sense, we must clarify that the rest of travel
for purposes other than work; they also experience reductions in time, but have not been
quantified in this work.
Territ orial impact of a development project
Finally, for each project group will do the following impact assessment: 1) changes the data
sections that are directly affected by the group. 2) Recalculate indicators (and its variation
percentage) of residents' socio-professional structure, activity structure, morphology. 3) Apply
the double-dimensional gravity model to determine the new matrix of t ravel, resulting from the
modification of both origins (residents) and destinations (jobs). 4) Calculated the new indicators
of accessibility (and percentage change) product of the modification of t he matrix of travel,
considering the matrix of travel times from the stage of evaluation. 5) Calculat ed the new values
of supply usage (and percentage change), using the hedonic model wit h the explanatory
variables changed in the preceding paragraphs.
The spatial hierarchy of values
The analysis of t he variation in spatial hierarchies downplayed the results of the implementation
of t he hedonic model (and especially its error) for the all territory as a whole. For this 10 scales
were defined in property values, which represent hierarchies.
The effect was evaluated is the change that produces the mat erialization of all transportation
projects considered for the future scenario (PDI) in relation to t he effect of the mat erialization of
all urban developments.
Each territ ory is affected by a project, change the property value, which can lead to a change in
a hierarchy to anot her (up or down), or to remain in the same hierarchy, despite the change in
value. While i n the previous paragraphs explained the change in value in percentage terms
(relative to the same value) is evaluated at this point the absolute change in value, but
relativizing its position to the territ ory. In this sense the situation may be a small percentage
change generates a change of hierarc hy, or that a large percentage change does not generate
hierarchy.
4. M AI N RES UL T S .
The results refer to the reports and calibration analysis of hedonic models, and the impacts
evaluated by projects in the different dimensions analyzed.
4.1 Calibration of hedonic models
Table 5 shows the ec onometric report has been simplified so as to indicate, for each real estate
product, the variables included in the final model (which were statistically significant at 95% ),
the sign of the coefficient, and finally model adjust (percentage of explanation of the real estate
2
value by the model-R -).
Table 2 Final report of the calibrated model
Variable inclosa
Model
Habitatge
Locals comercials
Locals comercials
Oficines
Venda
Venda
Lloguer
Lloguer
Valor total (euros) Valor total (euros) Valor unitari (euros/m2) Valor unitari (euros/m2)
Beta
Beta
Beta
Beta
Constant
10,57800
Característiques de l´oferta Superficie total construïda (habitatge)
0,01000
11,78000
6,67700
0,00390
0,24550
0,00400
0,39280
Indicador de qualitat integral (oficines)
0,06240
Superfície del local a nivell d'accés (comerç)
Localització entre 100 i 200 m de l´eix comercial principal
Resident
Factor socioeconòmic de nivell sota-alt
Factor socioeconòmic de nivell mig
-0,07778
-0,01074
Activitats econòmiques
Factor d´activitat industrial-no terciàri cualificat
Factor d´activitat terciàri banal
-0,03266
0,00144
Característiques del parc
edificat
Densitat d'habitatges 76 - 90 m2 (Concertat)
Densitat d'habitatges < 75 m2 (HPO)
Atractivitat i accesibilitat
Temps ponderat des d'origen al treball (TPO treball)
Densitat de viatges atrets de compres
Dependència espacial
Ajustament del model (R2 adj)
1,58800
-0,18910
-0,06700
-0,08430
-0,01412
-0,00002
-0,00002
-0,00002
-0,00294
0,00001
0,09252
0,02600
0,36340
0,18100
0,78
0,59
0,50
0,41
Source: Own CPSV
The model homes sales (t otal value) achie ved a 77% adjustment. The variables that mak e up
this model, in order to participate in the adjustment are: 1) tot al built area that adds value
(positive sign), 2) socioeconomic factor low -high, whose negative sign indicates that the
variable sum value (predominance high income levels), and subtracts value when the factor is
positive (prevalence of low income levels), 3) TP O to work, which have a negative sign
subtracts, with increasing travel time to work, home los es value. 4) Factor middle-and t ertiary
activity factor banal, as value added activity predominates tertiary banal, and when there
devalue residential administrative dominance, res pectively.
The model of retail shop sales (total value) reached a 59% adjustment. The variables that make
up this model are: 1) the local surface level of access adds value, 2) socioeconomic factor lowhigh, value added or subtracted as described in previous models, 3) the location within 100
meters a axis shopping that adds value, 4) socioeconomic factors mid-level, which subtracts
predominanc e, 5) Housing density between 76-90 m2 subtracts, 6) and the spatial dependence
of the values that provides the value.
The model of ret ail shop rental (tot al value) achieved a 50% adjustment. The variables that
make up this set are: 1) t he local surface level of access that adds value, 2) the spatial
dependence of the values that add value, 3) be located wit hin 100 meters of a axis shopping
that adds value , 4) and housing density of less than 75m2 which subtracts the value.
The model of office rental (unit value) reached a regular adjustment of 41%. The variables are
structured model be: 1) integral quality indicator of t he office offered that adds value (positive
sign), 2) spatial dependence of the value, this means that the values of other bids environment
influence its value unit, a positive (if the environment is high, the value is high, and vice versa),
3) socioeconomic factor low-high and density housing between 76-90 m2. The socioeconomic
factor low-high add / subtract value according to the predominance of high / low level of income.
The density of houses between 76 -90 m2 subtracts office space.
4.2 Impact of urban development
The impact evaluated of each group of urban developments concerning five major dimensions:
social and occupational structure of residents (POR), the structure of economic activities (LTL),
the residential types (type of housing density), accessibility of territory, and the value of potential
real estate deals. The results show that the groups of project s of higher percentage impact on
the social and occupational structure of the residents are La Marina - Zona Franca, the
Directional Centre P rat, and El Pla de P onent, with values above 200% of the base situation.
They are the largest projects in terms of residential and generate the greatest impact in the
receiving area (industrial or depopulated). Of t hese projects, only La Marina - Zona Franca
affects the structure in the mid-level factor, since it reduces the predominance of middle-income
class product of t he inclusion of a large proportion of public housing, while El Pla de Ponent
changes predominant structure by the high level by the fact that basically fits in the city of Gava,
historically working class. The group of 22 @ and Molins de Rei, change the structure of their
territories, the first by predominant low levels because it only considers public housing and lofts
that are discretionary in some heritage buildings, and the second produces the predominance of
upper classes as well as El Pla de Ponent transforms a little labor profile of the municipality to
include open-rate housing. The result also suggests that there is no direct relationship between
the absolute increase in population and the percentage increase and bet ween the percentage
increase and change in social and occupational structure. This change denotes territorial inertia
of the population, namely the change in magnitude is not necessarily significant percentage
change, nor change the social and occupational structure.
The results indicate, the impact on the structure of economic activity, approximately 50% of the
project groups generate job inc reas es over 100% of the base situation, and three of these
increases over 1000% over of the base situation (Cent re Directional Prat, Ca n 'Alema ny, and
Ciutat airport). It is important to clarify, in the c ase of 22 @, the change of structure refers to the
realization of 100% of the projects in relation to the 2006 situation in which some part of projects
already completed, and therefore already had c hanged the original structure (prior to 2000)
industrial markedly. The s ame remark applies to the other projects that by 2006 they were
already partially completed. Regarding the change in t he structure of activities, the frequent
changes are to the predominance of activities banal office and tertiary (commerce and
services). One can see that there are projects that increase the existing specializatión in the
basic situation of affected territ ories. This mixed-use projects, for example, the Ciutat Airport
which not only includes personal services and logistics activities, but also office space. Finally
one c an argue that in this dimension, the impact of projects is significantly higher, generating
most of them significant structural changes of economic ac tivities in the territory.
Regarding the impact on the residential types, the res ults show that few projects that provide
percentage increases over 100% (El Prat Directional Centre, La Marina - Zona Franca, and El
Pla de P onent). Most groups have a lower percentage inc reas e of 100%. The densities that are
most affected, with significant percentage increas es (over 100%) correspond to free housing.
The density of public housing (VP O) has lower percentage increases (less than 100% ), of which
highest percentages coincide with the largest total contributions from privat e homes.
The results on the impact of accessibility show that all the projects reduce travel time weighted
residence to work. This is very significant, and suggests that urban sprawl in the metropolitan
area of B arcelona, follows a pattern c haracterized by a polycentric, in which the old industrial
areas renewing with higher density levels in new businesses semi-peripheral as those at El
Prat, L'Hospitalet, Vilamoura, or the area B esos. These cent ers are generated the most
significant impacts on the overall reduction in travel time home-work. Finally the groups at
greatest impact on accessibility are those that generat e new cent ers of activity in the west
entrance of the central area of Barcelona. This occurs because of the good connectivity of these
points (concentration of transport services), and the high rate of generation of work travel
purposes, from the western zone, which are captured centralities above.
The effect of different groups of projects in the territorial baseline information, and accessibility
on the property value is a combination of all effects presented above, weighted by their hedonic
coefficients. To evaluate the effects of projects must be considered a "product property" specific
and assume that this product is offered in each of the studied census tracts (a situation that is
not real, but to appreciate the potential bid that would exist in all the territory of this product, and
its comparative hierarchy.) For the above products are c onsidered the following types (average
values of the samples used in the calibration stage): 1) residence: area of 82.36 m2, 2) offic e:
endnote quality 6.4 (scale 1 to 7) and 3) retail shop: area of 138.6 m2, with 98% of the s urface
at the entrance floor, and located within 100 meters of a main axis shopping. From each of
these products was evaluated bid value (predicted by the hedonic model) in the situation
without project, and after the implementation of the project. Ultimately not analyze the a bsolute
value of the offer (which is slanting by the model error), but rather the difference in value (in
percentage change) caused by the project, namely, the changes produc ed by the project on
accessibility, the resident social structure, the structure o f activities, and morphology
constructive.
The results suggest that the sales value of residence, 23 groups (77%) increase the value, the 3
groups (10%) have no effect (variation bet ween -1 and 1% of value) and 4 groups (13%)
decrease the value. The two groups most increase the value they do on a scale beyond 20%,
and correspond to the Directional Centre Prat, and La Mina - La Catalana. The group that
makes the value decreases by about 10%, and corresponds to Zona d'Activitats Logistiques.
The decrease in value produced by some groups of projects should be preferably the change in
economic activity rat es, though t hey do not succeed in changing the structure (change in the
principal factor), significantly increase the industrial activity, which by its sign, directly affects the
residential value. To this is added the increase in housing density concerted and HP O, as in the
case of projects and Molise Cornella de Rei, which also decrease the value. Regarding the
sales value of retail shop, shows that 18 groups (60%) increase value, 11 groups (37%) have
no effect, and only one group (3%) dec reases t he value. The groups do most increase in value
with a magnit ude above 20%, and correspond to the Pla de Ponent, and La Marina - Zona
Franc a. The only group to decrease the value (1. 4%) is Tres Turrons - Vall'd Hebron. In general
we can say that development projects increas es the value of sales of both residential and retail
shop. It is also necessary to remember that hedonic models sales value is those that show a
better position to specify parameters, and adjust the calibration sample. Therefore they are the
most reliable.
The results of t he rental value of office only two groups of projects (7% ) inc reas e value, 4
groups (14%) have no effect, and 24 groups (79%) decreas e value. The group further increas es
value, it is in a magnitude of 4%, and corresponds to Ca n Alemany. The group that decreas es
value, it does at a rate higher than 20%, and corres ponds to Prologis - Alcampo. As for the
rental value of retail shop, do not appear to increase the value groups, 20 groups (37%) have
no effect, and 10 groups (3% ) dec rease value. The group further inc reas es value, it is in a
magnitude of 7.5%, and corresponds to La Marina - Zona Franca. In general we can say that
the effect of the groups of urban projects in the rental value, both in offices and retail shop, is
not to alter the value, or decrease in low magnitude. At this point it should be noted that the
rental hedonic models are those that are least able to specify paramet ers , and adjustment.
Therefore the results presented must be considered with great caution.
Finally, it presents findings of the Impact on the hierarchical structure of sales value. In this
analysis, values are considered more appropriate adjustment determin ed by the hedonic
models (higher rates) and are: 1) the sale of residences and 2) the sale of retail shop. The
hierarchies considered c orrespond to calculate t he deciles of sales values predicted by the
hedonic models for the bas e situation (2007). Figure 3 shows the s patial structure of the
hierarchies of the sale price of homes in the base year.
Figure 3 Hierarchy and effect on the sales value residential
Source: Own CPSV
Trans port projects (PDI) only inc reases of hierarchy in a single category (namely moving from
one cat egory to the next), but in a marginal amount (only 0.8% of the census tracts). The urban
development changes occur in approximately 13% of the s ections analyzed, of which about 5%
of them decreases the hierarchy (mainly low one category), and 8% up the hierarchy (preferably
up to a higher category). While on the stage of urban development, 87% of the census tracts
are not affected in the values of residential offer for sale, you can suggest a slight tendency fo r
emergence of new centers of high value, which is not generate a significant difference in the
structure of the base year.
In the case of the sales value of retail s hop, the hedonic model does not include accessibility
variable, so that transportation projects do not generate effects on these values. Figure 4 shows
the territorial hierarchy of values of retail shop sales.
Figure 4 Hierarchy and effect on the sales value of retail shop.
Hierarchy 2007
Effect urban developments
Source: Own CPSV
The urban development changes occur in approximately 7% of the census tracts studied, of
which about 2% of them decrease their hierarc hy and a 5% increase in hierarchy, also achieved
significant increases. The scenario of urban developments in respect of the sales value of retail
shop, does not suggest a clear trend of emergence of new centers of high value, as in t he case
of residential sales value.
5. C O NCL USI O NS
This research attempts to understand what is the impact of large urban development and
transport infrastructure projects planned for the area east and west of the central conurbation of
Barcelona, on accessibility, land uses, and values property.
Through the integrat ed use of transport models and hedonic pricing models we attempt to
simulate the direct impacts in the area of implement ation of a specific development project, and
also the indirect impact that occurs in the system of interaction of the metropolitan area
Barcelona.
The different projects grouped based on their location and logic of design, produce different
effects on t he dimensions analyzed. There are situations in which t he contribution of the project
is significantly different from the territory which it hosts, so the property value does not vary
significantly, or vice versa, that is, small changes in the bas e situation creates significant
changes in property values. Thus we can conclude that there are situations territorial "buffer" or
"amplifying" the effect on property values and that these characteristics depend on the initial
situation of the territ ory for the key variables identified in the hedonic models.
Impacts on accessibility dimension beyond the territories receiving the different urban
development, creating areas of affectation that depend on net work configuration, the current
system of transportation services, and its future shape. It is for this reason that the urban
development located at points of entry to the central area of Barcelona (Sagrera, Sant a Adrià,
Badalona, L'Hospitalet de Llobregat, etc.) generate effects of highest catch of potential
commute, and therefore, reduced travel times.
As for the groups of projects that reduce property value offer, it is concluded to be mainly to
changes in s ocial and occupational structure of residents a nd t he activities (jobs), which mainly
affects the value of home s ales , and secondly the sales value of ret ail shop. So can you
conclude about the great importance which holds , the social structure dimension of the territory
in the spatial structure of property values.
The spatial structures of hierarchies of value of residential and commercial sales, suggest that
the effect of urban development as a whole, is more significant than t he effect produced by the
implementation of transport projects (PDI) planned for the study area. Moreover, the effect
generated by urban development, in the spatial structure of hierarchies of sales value, shows a
slight tendency of new centers emerging high value (preferably in nursing), which does not
stand out significantly from the hierarchies the base situation. Finally the impact of urban
development in the reduction in travel times although not large in size, generated greater equity
in access conditions.
The methodology used in the area assigned to it, opens up new poss ible line of analysis,
despite its deficiency and limits of application. The possible lines are: 1) determine the level of
damping or reactivity in property values, in different areas within the study area, 2) determine a
profile "optimum" of urban development, on the types of activity and intensity thereof, so as to
generate desired effects of appreciation in different areas and environments, 3) Identify
strategic partnerships of trans port projects and urban development, to revalue depressed areas,
4) identify variables territorial "control" of speculative processes real estate value and territorial
factors that hinder the emergence of real estate supply in different territories, and 5) validate
the model results with observed data.
R EF ERENC ES
Autoritat del Transport Metropolità, Pla Director de Mobilitat, Barcelona, (2007). Online.
http://www.atm.cat/cat/apartado3/ap3_04.htm
Banco Interamericano de Desarrollo, Manual de E valu ación Económica de P royectos de
Trans port e (2006), Washington D.C., pp. 188. Online.
http://idbdocs.iadb.org/ wsdocs/getdocument.aspx?doc num=1037792
HUNT, J.D. (2005). Current Operational Urban Land-use-transport Modeling Framework s: a
review. Transport Reviews, Volume 25, Number 3, pp. 329-376.
DE SOUSA, C.; WU, C. and WESPHAL, L. (2009) Assessing the Effect of Publicly Assisted
Brownfield Redevelopment on Surrounding Property V alues . Economic Development Quarterly.
Number 23.
MCMILLE N, D. (2008). Changes in t he distribution of house prices over time: Structural
characteristics, neighborhood, or coefficients? Journal of Urban Economics, Number 64, pp
573–589.
BRASINGTONA, D.; HITED, D. (2005) Demand for environmental quality: a spatial hedonic
analysis. Regional Science and Urban Economics, Number 35, pp 57– 82.
LEE, B.; CHUNG, E. y K IM, Y. (2005) Dwelling Age, Redevelopment, and Housing Prices: The
Case of A partment Complexes in S eoul. Journal of Real Estate Finance and Economics,
Number 30:1, pp.55–80
SONG, Y. KNAAP, G.-J. (2004). Measuring the effects of mixed land uses on housing values,
Regional Science and Urban Economics, Number 34, pp. 663–680.
CERVE RO, R and DUNCAN, M (2004). Neighbourhood Composition and Residential Land
Prices: Does Exclusion Raise or Lower Values? Urban Studies, Volume 41, Number 2,
February, pp. 299–315.
SIMANS G.; MACDONALD, L.; MACP HERSON, D. and NORMA N, E. (2006) The Value of
Housing Characteristics: A Meta Analysis. J Real Estate Finan Econ, Number 33, pp. 215–240.
SIMONS, R.; QUE RCIA, R. and MERIC,I. (1998). The Value Impact of New Residential
Construction and Neighborhood Disinvestment on Resid ential Sales Price. Journal of real
estate research. Volume 15, Number 1/2, pp. 147–161
GIBBONS, S. AND MA CHIN, S. (2008) Valuing school quality, better transport and lower crime:
evidence from house prices, Oxford Review of Economic P olicy, Volume 24, Number 1, pp.99–
119.
Descargar